Written by 7:30 am AI, Latest news

**Enhanced AI Workload Support with OpenStack**

OpenStack allows enterprises to manage their own AWS-like private clouds on-premises. Even after 29…

OpenStack enables enterprises to oversee their own on-premises private clouds similar to AWS. Despite 29 iterations, it remains one of the most active open-source projects globally. Recently, the OpenInfra Foundation, which guides the project, introduced the latest version, named ‘Caracal,’ focusing on enhanced features for hosting AI and high-performance computing (HPC) workloads.

The primary users of OpenStack are large corporations such as Walmart or NTT, a major telco company. Currently, most enterprises are exploring ways to deploy their AI models while ensuring data security, often opting for full stack control.

OpenInfra Foundation CEOO Mark Collier

Mark Collier, the COO of the OpenInfra Foundation, highlighted the impending massive investment wave in data center infrastructure, as noted by Nvidia CEO Jensen Huang. A substantial portion of this investment will target private deployments, necessitating a software layer like OpenStack for management.

OpenStack now stands out as a comprehensive alternative to VMware’s offerings, especially with VMware users seeking options due to concerns following its sale to Broadcom. Jonathan Bryce, the executive director of the OpenInfra Foundation, mentioned an increasing number of companies reconsidering OpenStack amidst the VMware acquisition and licensing changes.

Image Credits: Frederic Lardinois/TechCrunch

OpenStack has witnessed significant growth in the Asia-Pacific region, with recent adoption by Okestro, a South Korean cloud provider focusing on AI. Additionally, Europe, with its stringent data sovereignty regulations, has emerged as a growth market. Notably, the UK’s Dawn AI supercomputer operates on OpenStack.

Mark Collier expressed optimism about the surge in open-source infrastructure adoption, particularly for projects like OpenStack and Kata Containers, indicating a promising future for infrastructure upgrades.

The latest OpenStack release introduces vGPU live migrations in Nova, facilitating the seamless transfer of GPU workloads between physical servers. Security enhancements, such as rule-based access control for core services like Ironic, and networking updates to support HPC workloads, have also been implemented.

The recent update marks OpenStack’s transition to the ‘Skip Level Upgrade Release Process’ (SLURP) introduced a year ago. This approach streamlines the upgrade process, offering a stable and more manageable cadence for enterprises. The project has evolved into a mature system with a sustainable ecosystem, positioning itself well to leverage the AI boom.

In conclusion, OpenStack’s journey has seen fluctuations in perception, but it now stands as a reliable system with ample opportunities to capitalize on the evolving landscape, particularly in the realm of AI.

Visited 2 times, 1 visit(s) today
Tags: , Last modified: April 5, 2024
Close Search Window
Close